PhaseMark: A Post-hoc, Optimization-Free Watermarking of AI-generated Images in the Latent Frequency Domain

2026-01-19

AI Summary

This work addresses the inefficiency of existing post-processing watermarking methods for AI-generated images, which typically rely on iterative optimization or inversion procedures and are thus impractical for real-world deployment. To overcome this limitation, the authors propose PhaseMark, a novel framework that embeds watermarks without optimization by leveraging phase modulation in the latent frequency domain of a variational autoencoder (VAE), requiring only a single forward pass. PhaseMark introduces four distinct phase modulation variants that achieve state-of-the-art robustness against strong attacks such as regeneration while preserving high visual fidelity. Notably, the method accelerates watermark embedding by several orders of magnitude compared to optimization-based approaches, establishing a new paradigm for efficient and robust watermarking of AI-generated imagery.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Efficient ML / Green AI
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The proliferation of hyper-realistic images from Latent Diffusion Models (LDMs) demands robust watermarking, yet existing post-hoc methods are prohibitively slow due to iterative optimization or inversion processes. We introduce PhaseMark, a single-shot, optimization-free framework that directly modulates the phase in the VAE latent frequency domain. This approach makes PhaseMark thousands of times faster than optimization-based techniques while achieving state-of-the-art resilience against severe attacks, including regeneration, without degrading image quality. We analyze four modulation variants, revealing a clear performance-quality trade-off. PhaseMark demonstrates a new paradigm where efficient, resilient watermarking is achieved by exploiting intrinsic latent properties.

Problem

• watermarking
• AI-generated images
• Latent Diffusion Models
• post-hoc
• optimization-free

Innovation

• PhaseMark
• latent frequency domain
• optimization-free watermarking
• phase modulation
• diffusion model watermarking